Indiana Temperature data

Saturday is when I look at State Temperature data, and after looking at Illinois last week, I thought Indiana might be a reasonable next choice, moving East after the sojourn out West that started the series. Having found, as with Missouri, little correlation of temperature with time, this week’s hypothesis is that the Central States are less sensitive to changes in temperature over time, than the rest of the country. So let’s see if it is true. (For those new to this I try and pose a hypothesis at the start of the post, and then, writing the post as I run the calculations, find out if it is true. Some are, some aren’t). (And for those who have been reading these for a while there was a significant surprise this week – see if you spot it – grin - it proves I write these as I do them).

It is a little irritating to note that the temperature values I am using are not actually raw data, but as Dr. Hansen noted recently, the USHCN data is already adjusted for urban heat island effects.

We use the unadjusted version of GHCN. However, note that a subset of GHCN, the United States Historical Climatology Network (USHCN), has been adjusted via a homogenization intended to remove urban warming and other artifacts [Karl et al., 1990; Peterson and Vose, 1997].

It is an interesting note that this current study shows that the effect of population size on the data remains present, in many states, and we will now see if it holds true in Indiana. But at some stage I suppose I will (once I get sufficient states under my belt to be credible) have to see about the raw data issue.


So starting off for Indiana with the USHCN Network for that state, there are 36 stations so there will be a slight pause while I create the spreadsheet and import the data. Oh, and our first surprise of the day. The web site for the data has been reconfigured, so that the new page (this is going for Angola, which is the second USHCN site in Indiana) carries a number of options, rather than just the one set of adjusted data that I had been using until now:


So the question is whether we want the fully adjusted data, the raw data, or the time of observation adjusted data. Well I should run a series of tests to see what the difference between the sets turns out to be – but 36 stations is a lot of data to load, so I will probably do that at a later date for a smaller state, and for now use the data set that is merely adjusted for time of observation bias (The TOBS data set). So let’s get those into the spreadsheet.

And to remind you how I am doing that, you select the middle file option to download:


Click on that brings up this screen:


Select the average mean temperature (TOBS) and click the submit button at the bottom.

This brings up the download page:


Click on the blue line, and this downloads the file, which I then save as Angola Ann Temps on my Desktop, since this makes it easier to then find, and download the data into the spreadsheet. (See Missouri file for how I do that. The states I have covered already are in a list at the very bottom of the right-hand column on the site).

Well, having done that I discover, using Angola for the review, that the RAW and TOBS data sets have years missing, and that it is only the fully adjusted file that has the data for every year. Tsk! Well, had I more time today I would do a little adjusting of my own, but since I don’t, I will set that to one side for today, and go back to using the adjusted mean values for today. (Too many other things to do). Well at least I caught this early).

But just to give an example of the differences, here are the RAW, TOBS and adjusted data for the first 12 years of each data set. (There is then a gap of several years in the RAW and TOBS files – though not the ADJ file).


You will notice that the RAW and TOBS files have several temperatures in the 50’s, while the ADJ file does not. Hmm! The average RAW data temperature is 48.7 deg; the average TOBS temperature is 49.4 degrees: the average ADJ temperature is 46.3 deg. Well how does that compare with more recent temperatures – let’s look at the most recent 12 years:


You can see that the ADJ values (the ones that are used by GISS) are much closer to the RAW and TOBS values, with the average RAW temperature being 48.3 deg F, the TOBS value being 48.8 def F, and the ADJ data being 48.7 deg F. Now if I were tempted to be suspicious it would seem a little odd that the adjustment to get the used data lowers the historic temperatures while leaving current values more alone. But to substantiate that I would need to look at a lot more data than I have time to examine today, but it does cast some slight suspicion on the data as I move forward, for now using the Adjusted data.

So now, we move back to getting the data, and that means re-acquiring Anderson as an Adjusted Mean temperature. And so we get the data from the USHCN site, now carefully marking it as adjusted, and go to get the GISS data. There are four GISS stations in Indiana, according to Chiefio and these are in Evansville, Indianapolis, Fort Wayne and South Bend.

There is no data for 1895 or 1896 for Evansville and Fort Wayne, and South Bend data does not start until 1944. In the past I have interpolated data, or used adjacent sites to generate those numbers. But this time I will just average the value for the GISS stations based on the number of stations in that one year. Doing that clearly shows that the number is small enough, that each has a significant impact. Consider the difference between the GISS data and the USHCN data:


The change when the fourth GISS station is added is clear. Looking at the two segments of the curve, however, it is also clear that there is an increasing difference between the GISS stations and the USHCN stations, in both. This would indicate (and remember that the cities GISS chose are some of the largest in the state) that they are warming more than the average for the state.

For the state as a whole (now averaging all stations, for each year) the influence of the lack of one station becomes much less. But, as predicted by the hypothesis at the beginning of the post, there has not been a significant (i.e. R-squared greater than 0.05) increase in temperature over the 110 years.


The state is fairly flat, and longer North-South so that there is enough scatter for a good correlation with Latitude, and as with every other state, that is evident from the data.


And since there aren’t any strong geological features East-West I wasn’t looking for a strong correlation with longitude, and there isn’t one:


However, even though the state elevation only fluctuates over a couple of hundred meters, there is a strong correlation with elevation.


The other two graphs that I usually plot are the standard deviation of the data over time (to see if Anthony Watts concern about station deterioration is true – if so SD should perhaps increase over time.


And it does appear as though, within the past decade or two, that quality is beginning to suffer.

And so the question is, with the data for population being “factored out” of the data by the USHCN, is there a population trend. Bear in mind that the states where this was evident were states where there were a lot of very small communities. In Indiana the communities with stations are increasingly larger.

Hmm! As with Illinois last week it appears that there is no longer a correlation with population. Maybe the USHCN correction is a good one? (Perhaps, since the tabulation is new, I should go back and see if the earlier state data is still the same?)


Well, more interesting results. The story seems to be getting curiouser and curiouser the more we look. And we aren’t even half-way done yet!

Maine Temperature information

This is a part of a continuing look at state temperatures, and, since time is a bit short today, and since this was written in Maine*, I thought I would look at the temperatures there, rather than doing California, which with over 50 stations will take a little longer. Maine also provides an interesting contrast to Nevada, having some of the coldest temperatures in the contiguous United States, rather than some of the hottest. The process is the same as that I started with, and so I will be comparing the dozen USHCN stations in Maine with the single station that GISS uses. I wandered around trying to check that Caribou was the valid GISS site, as Chiefio listed. Along the way I bumped into the US Climate Reference Network (USCRN).

When fully deployed, the USCRN will consist of approximately one hundred stations nationwide at locations selected to capture both the national and regional climate trends and variations for temperature and precipitation.

The reason that I question the GISS use of Caribou, is the same as that which gave me problems when looking at Nevada. The temperature data from the site does not go back to 1895, as the USHCN data does, but in this case only starts in 1948, making a true comparison over the period of record impossible.

So, again being curious, I looked for a listing of the USCRN stations, which turn out to be different from the one provided by Chiefio. There are two stations on this USCRN list for Maine, one at Limestone (20020920) and one at Old Town (20020913). And this gets even more confusing since the data for Old Town from GISS only run from 1949 to 1972, and that for Limestone doesn’t apparently exist (the town has 1,415 people in it.) Perhaps it is a new station that is being installed? Maybe I should find out what is going on? But the record hasn’t been good on folk getting a response from those responsible (see the Climategate e-mails). So, given the confusion, I think that I will just stick with the Caribou station for now, since it does, at least have a significant length of time over which there is a record, that can be compared with the USHCN data.

However, by using this station I do invalidate one of the hypotheses that has been evolving from data from the last five states. For this is the first state where the GISS station temperature is lower than that of the average of the USHCN stations. Plotting the difference between the two, this really doesn’t appear to have changed over the time interval where there are records for both:

In Maine there has been a steady increase in state temperature over the past 114 years:


Now I had a second hypothesis that I was going to test, and this was that the influence of the sea would be evident in a warmer temperature near the coast, which dissipated as the stations moved west, and both away from the coast and higher. Well that one was wrong, at least in regard to it becoming cooler as the stations were further west and higher:


However there is a caveat to this result, in that the coast of Maine runs at an angle of about 45 deg along the bottom of the state, so it might be that we need to filter out the effects of latitude and altitude before having a closer look. Also, as I will note again later, there are only a limited number of stations in the state and two of them are sensibly on the coast.

The linear correlation with Latitude that has been detected for all the states so far, has persisted:


As has the linear inverse temperature correlation with Elevation above the sea, even though, in this case, the heights are not that great:


The change in standard deviation over time follows the pattern that we have seen in many of the states, starting out rising, and then, as automated temperature recording was introduced, reducing.


And so I now adjust the data by adjusting each station data to the latitude for the Center of the State, given as 45.250556, -45.250556.

Then I tried adjusting the data to the value if the station was located at the average height above sea level for the state of Maine, which is 600ft. And discover that this doesn’t work for all stations, since those that are below 100ft above the ocean tend to be dramatically overcorrected.

This comes back to the point that there are only a few stations in the state, and that there are only two that are that close to sea level, and we don’t have enough data at this time to draw a conclusion. Which means finding a state with more stations at the coast. . . . .

In the meantime to round out the usual set with a variation, here is the mid-state average temperature (i.e. adjusting each station for latitude as though it were at mid-state, and then taking the average – without Caribou – though it made very little difference). (To do this I subtract the constant term from each temperature based on the temp v latitude graph, then I multiply by the latitude of the center of the state, divide by the latitude of the station, then add the constant term back in).


And then to look at the effect of population, though again using the temperatures adjusted for latitude:


And we can still see that stronger sensitivity to population at small town sizes, that has been consistent across the states examined to date – but then they have all had relatively small population sizes.

* A series of problems that began with problems in getting onto the internet at the hotel, through having to change the router on our system when we got home, ends up with this also being completed on a new computer a week after the vast majority of the post was first written, sigh!

Utah Temperatures and Colorado UHI

This is now the fourth in an originally unplanned series looking at the differences between GISS and USHCN average temperatures and some trends in the data that are not necessarily obvious when one only looks at global data. Tracking Westward from Colorado, Utah has the next set of temperature data that I will look at, and I will also comment on Anthony Watts note on Colorado UHI , which he posted since my post on that state.

Utah has a different topography in that while in Colorado as we head West the land rises, in Utah that is no longer the case. So for the hypothesis of the week I will assume that while there will be a correlation with elevation, that there will be none with longitude. Let’s see if I’m correct. And while there was little correlation with latitude in Colorado, because of the strong effect of elevation and the wide range it covered, I am going to hypothesize that we will see a significant effect of latitude again, even though the elevation is considerably higher.

As usual I am writing this as I carry out the tabulation, and so I begin, following the initial procedure, by going to the USHCN web site and, yikes, there are 40 weather stations listed for Utah. OK, there will be a slight pause while I enter all these into the data table . . . . 40 stations, mutter, mutter).


Hmm! And we have the same problem of missing data that showed up in Colorado, but with more stations. For some stations there is no data in specific years before 1900. Well I am going to apply the same procedure as I did for Colorado and calculate an assumed temperature for those sites. First is Blanding, for which there is no temperature for 1896. Blanding is, on average at 51.15 deg, 2.8 deg warmer than the rest of the state. If that year the average temp was 47.5 deg, then Blanding should be at 50.3 degrees. If the average temperature for 1896 was, at 47.5 degrees, 0.8 degrees below the average annual temp, and the annual temp for Blanding is 51.15 deg, then the temp in 1896 would be 50.35 deg. Taking the average of these gives 50.33 deg, which is what I insert.

I follow the same procedure to calculate the temperatures for Bluff in 1895, 1896, 1896, and 1898; Farmington in 1895; Morgan Power in 1895; Snake Creek in 1895; (if you get the table these values are marked in red.)

Again checking on Chiefio’s list of the GISS stations. I find only one station out of the stations in Utah, that has survived the cut is the one in Salt Lake City. So I go to get the GISS data for that station.

Adjusting the tabulation to calculate averages for the different numbers of USHCN and GISS stations reveals that the GISS station reads 4 degrees higher than the state average from the USHCN stations.

Looking for the populations of the different cities and towns, Deseret is now apparently preferably called Delta. Modena, UT is also not on the city-data file, and so I got the population of 35 from Century 21. The population for Snake River Powerhouse also requires a little detective work, and two communities have merged to form Midway, and so I will use that number. And for Zion National Park I used Springdale.

And so all the data is in the table and what do we find? First of all, looking at the primary hypothesis of the day:


There is realistically no correlation here, so the first hypothesis is apparently valid.


As with the Colorado data, this again suggests that the trend that we saw with the Kansas data in regard to longitude was really a reflection of the increasing height above sea level. And at the high elevations, looking at latitude, there is a trend:


Not as strong as elevation, given the widely carrying heights of the stations around the state, but it is still there.

In regard to the population distribution, Utah is a state with a large number of small communities , which also have weather stations, so I shrunk the scale of the plot shown to below 10,000 population:


And the log correlation still shows.

In terms of the standard deviation plot that I have calculated for all the states so far:


There has been a slight increase in scatter over the years, (which would validate Anthony Watts observation on station quality) though it is not at a significant level (which I have set at an r-squared of 0.05).
However Utah is a state that has seen a warming over the years:


But the growth of the small communities, relative to the single GISS station in Utah, and the sensitivity of temperature more to the growth of small communities, means that the difference between the GISS and USHCN averages has declined over the years, due to the makeup of the USHCN average.


Which brings me back to Anthony Watts recent post on Colorado data. What he and Stephen Goddard have done is to look at the relative population growths of Boulder, and Ft Collins in Colorado.

Growth in relative populations of Boulder and Ft Collins in Colorado

Now, taking the data that I had discussed last Saturday, it was fairly easy to pull out the data for Ft Collins and Boulder and find the difference, and plot it against time:



I have taken it back a little further than WUMT, since I had the data, but the two curves are otherwise the same. And it does show the effect of the Urban Heat Island, from the point of view that the most significant change between the two towns has been their relative growth.

And so it continues, perhaps next week we will look at Nevada?

Temperatures in the Mountains - the Colorado data

Today I am going to look at the temperature data for Colorado, since last week we looked at the data from Kansas, and the correlation with longitude was growing, but could not just be explained by a change in height. The procedure starts off the same as that for the initial post, in that I am going to take data from the US Historical Climatology Network compare it with the GISS data for Colorado, (which I reference later) and then see if I can draw some conclusions from the data. I want to have a hypothesis to test against the data, and since the land rises in the West, and it gets cooler, the hypothesis this week is that there is a relationship that we can derive between average temperature and height above sea level. There is a subsidiary corollary to this, which is that this will explain the changes in temperature with longitude.

So to begin we find the data for the weather stations for Colorado. There are 24 of these, and so I begin by generating the table that I described in the first post of this series. This lists the site with its geographic location and leaves a space for the population. Under that I put the average annual temperature data for that site from 1895 to 2008. (and since I am writing this as I do it, there will now be a pause . . . .)

And there is a slight problem – in 1896 there is no temperature data for Telluride. Hmm! How to handle this? Well looking at the averages that for 1896 is on average 0.6 deg above that for the stations in the rest of the state. Telluride, on average is 9.54 degrees below the average for the state. So if I take the average for the state for 1896 and subtract 9.54 I get 37.1 degrees. If I take the average for Telluride and add 0.6 degrees I get 37.17 degrees. So it seems fair to insert a temperature of 37.135 degrees for Telluride for 1896 – which I did to complete the initial table.

Now to get the data from GISS, and here there is another surprise – I have checked the list that Chiefio gives 3 times and still can only find Grand Junction as a site in Colorado, even though the GISS site lists some 33 sites with data, of which a fair number also appear on the USHCN list. So I will just use this as the sole GISS site for Colorado. Getting the data from GISS and correcting it for scale (from Centigrade to Fahrenheit), then I run the table, and the next surprise of the evening. The GISS site is, on average over the 114 year period 6.65 degrees warmer than the state average. Plotting the average difference over that time interval


Now there is only one GISS station, but this is rather a large amount warmer for the state according to GISS than the state average would suggest. It has, however, been declining slightly, but steadily over the years. As for the state temperature as a whole, that has been increasing, though the pattern is a bit strange.


So now I go to enter the population, and not being a Colorado native I had been wondering where the Denver data was, and apparently it is hiding behind the Cheesman file. (That being a suburb of Denver apparently with a Park). So do I use the suburb population 8,201 or that of Denver itself – 598,707? Given that the area has a high population density, I am going to use the Denver number.

And then there is the problem of finding Hermit, CO – fortunately the weather station information includes the co-ordinates (since I didn’t have a great deal of success with a Google search) and this allows me to use Google Earth to go to the co-ordinates and find that it is at Hermit Lakes, which is in Creede, population 377. Putting all those together, there really are a lot of small communities in Colorado, so I’ll use a normal (rather than a log) plot of the information.


And with this having a lot of data at the smaller end of the scale (and recognizing that we have yet to go to a state with large populations) there is some correlation to a log relationship. And changes in small populations could have a greater significant effect. (Which is what I had said earlier, and which is now being also being written about by Professor Roy Spencer.

Interesting where there weren’t any large mountains (i.e. Missouri) there was a good correlation with latitude, but in Colorado that is not as evident:


And instead, where further East there was no correlation with longitude, in Colorado it is very pronounced:


So the question is, can this all be explained by the changes due to height above sea level, or elevation?


I don’t think that there can be any doubt of the correlation. So now we have two – where there is not a large change in elevation (Missouri) there is a strong correlation with latitude, but when the stations are at a higher altitude, then there is a strong correlation with elevation. And the starting hypothesis, this time, is seen to be correct.

One wonders how it would be if we normalized the data to account for both elevation and latitude. Given the lateness of the hour I won’t do that tonight, but given that GISS gives the locations for the centers of the states, maybe I will adjust the data to that location based on the linear relationships and see what that produces, and how much variation it takes out. But I’ll do that later in the week (we have been beset by server problems today, and I would like to get this out before I get hit with another).

Is global warming regional - as in the MWP?

I was looking over Anthony Watts’ web site, and noted a picture that he had posted that shows the large number of stations around the United States that were showing record temperatures.

Places in the United States showing record temperatures (From WUWT)

What struck me was that, just as one could argue that this showed how cold we were getting, those that argue that carbon dioxide is raising global temperatures could as easily point to the points in the Western part of the country as evidence for their argument.

The problem of trying to determine which arguments are more accurate is made more difficult where there is this isolation of data, so that those seeking information are only presented with that which supports one side – so cheers to Anthony for showing both.

But individual data points, while interesting, don’t really reflect the sort of long-term changes that are predicted (or not) by various climate models. There is a considerable question over the accuracy and methodology used to derive an average measure of the global temperature, with satellite data, which was supposed to be more accurate, having its own limitations, in measuring ocean temperatures for example.


If one looks at the average temperature for the contiguous United States, for example,

Average US temperatures since 1880 (GISS )

One can see that the average for the country has been tending downwards over the last decade, and has not differed that much from the temperatures of the 1930’s (which still has the record for the hottest year). It is hard to argue for AGW based on this graph alone.

Yet if once compares the data from the two hemispheres, it is the Northern Hemisphere that is warming much faster than the Southern.

Hemispheric temp changes (GISS)

So if the US landmass isn’t warming, and the Southern Hemisphere isn’t currently doing much either where is all the global warming coming from? A look at the current temperature anomalies shows that it is the Arctic and Europe that is warming the most:

Global temperature anomalies (GISS)

But it is interesting that the actual anomaly is only 0.54 degrees, since, as I have noted before, the predictive values that Dr Hansen gave in the paper that is the most cited as authoritative on this predicted that by now the temperature would be between 1.0 and 1.2 degrees higher than the datum, and that only if there was dramatic reduction in greenhouse gases would we get to 0.65 degrees by now.

Now that is still a it higher than the 0.42 degrees – which Anthony is reporting and which is, I presume, the daily difference, but as he notes, the carbon dioxide levels are at 388 ppm, and if the impact was a severe as has been modeled it should be much higher (though the GISS graph above does show that it is heading backup).

But the other thing to comment on is that back when those of us who bring up the Medieval Warming Period (MWP) first did so, the response of the AGW proponents was that we were discussing something that was only true of Europe and the Arctic. Now that we are seeing a pattern that shows a regionalization of the global warming pattern I wonder if that argument can’t be reversed?

It is only by openly discussing these changes and their implications that true scientific understanding can be achieved, not by hiding, or adjusting data in less than transparent ways. Unfortunately scientific debate takes time and evidence only accumulates slowly. How long, for example must the current apparent cooling persist, (or the accelerated increase in temperature not occur) before the model predictions are discredited? Yet without that debate we risk the frittering away of resources on valueless measures that will have no impact on the future.